There has been a resurgence of interest in dynamic factor models for use by policy advisors. Dynamic factor methods can be used to incorporate a wide range of economic information when forecasting or measuring economic shocks. This article introduces dynamic factor models that underlie the data-rich methods and also tests whether the data-rich models can help a benchmark autoregressive model forecast alternative measures of inflation and real economic activity at horizons of 3, 12, and 24 months ahead. The authors find that, over the past decade, the data-rich models significantly
improve the forecasts for a variety of real output and inflation indicators. For all the series that they examine, the authors find that the data-rich models become more useful when forecasting over longer horizons. The exception is the unemployment rate, where the principal components provide significant forecasting information at all horizons.